ABSTRACT: Information technology teams frequently manage projects, issues, approvals, requirements, risks, vendors, performance indicators, and operational records across separate tools. This paper presents the design and implementation of an integrated web-based IT management dashboard that connects operational project management with machine-learning analytics and decision support. The system uses a React/Vite frontend, FastAPI application services, SQLAlchemy data access, and PostgreSQL persistent storage. It provides modules for projects, issues, approvals, performance, requirements, vendors, risks, records, daily updates, and AI analytics. The AI layer integrates K-Means clustering, Linear Regression, Logistic Regression, and Random Forest using a shared project feature representation based on progress, issue count, high-severity issue count, and status score. The system also includes role-aware workflows, management approvals, requirement justification, and audit logging. Preliminary functional verification shows that the architecture can execute clustering and regression services when input conditions are satisfied and can explicitly report insufficient class diversity or training data for classification models. The work is presented as a functional research prototype rather than as evidence of production-level predictive accuracy. Because the current operational dataset is small, statistically reliable model validation requires a larger historical dataset containing repeated project observations
KEYWORDS: IT dashboard; project management; machine learning; K-Means; Linear Regression; Logistic Regression; Random Forest; risk analysis; decision support
LIPL IT DASHBOARD WITH AI-BASED ANALYTICS AND DECISION SUPPORT SYSTEM
MR. SAIRAJ SUBHASH MHATRE, MRS. ANJALI DANDER